Episode 393 The Mountains of the Moon

Today we’re going to the Mountains of the Moon – but not those on the moon itself. We’re going to central Africa.

There isn’t really a mountain range specifically named the Mountains of the Moon. The ancients, from Egyptians to Greeks, imagined or heard rumor of a mountain range in east-central Africa that was the source of the river Nile. In the 18^th and 19^th centuries, explorations of the upper Nile found the sources of the Blue Nile, White Nile, and Victoria Nile and identified the Mountains of the Moon with peaks in Ethiopia as well as 1500 kilometers away in what is now Uganda. Today, the range most closely identified with the Mountains of the Moon is the Rwenzori Mountains at the common corner of Uganda, the Democratic Republic of Congo, and Rwanda.

This location is within the western branch of the East African Rift system, an 8,000-kilometer-long break in the earth’s crust that’s in the slow process of tearing a long strip of eastern Africa away from the main continent. We talked about it in the episode for December 16, 2014.

The long linear rifts in east Africa are grabens, narrow down-faulted troughs that result from the pulling apart and breaking of the continental crust. The rifts are famously filled in places by long, linear rift lakes including Tanganyika, Malawi, Turkana, and many smaller lakes.

Virunga Mountains (2007 false-color Landsat image, annotated by Per Andersson : Source)

When rifting breaks the continental crust, pressure can be released at depth so that the hot material there can melt and rise to the surface as volcanoes. In the Rwenzori, that’s exactly what has happened. The Virunga volcanoes, a bit redundant since the name Virunga comes from a word meaning volcanoes, dominate the Rwenzori, with at least eight peaks over 10,000 feet high, and two that approach or exceed 4,500 meters, 15,000 feet above sea level. They rise dramatically above the floors of the adjacent valleys and lakes which lie about 1400 meters above sea level.

These are active volcanoes, although several would be classified as dormant, since their last dated eruptions were on the order of 100,000 to a half-million years ago. But two, Nyiragongo and Nyamuragira, have erupted as recently as 2002, when lava from Nyiragongo covered part of the airport runway at the town of Goma, and in 2011 with continuing lava lake activity. Nyiragongo has erupted at least 34 times since 1882. The volcanic rocks of these and the older volcanoes fill the rift enough that the flow of rivers and positions of lakes have changed over geologic time.

Lake Kivu, the rift lake just south of the volcanoes, once drained north to Lake Edward and ultimately to the Nile River, but the volcanism blocked the outlet and now Lake Kivu drains southward into Lake Tanganyika. Local legends, recounted by Dorothy Vitaliano in her book on Geomythology, Legends of the Earth (Indiana University Press, 1973), tell the story of demigods who lived in the various Virunga volcanoes. As demigods do, these guys had frequent arguments and battles, which are probably the folklore equivalent of actual volcanic eruptions. The stories accurately reflect – whether through observation or happenstance – the east to west migration of volcanic activity in the range.

The gases associated with the volcanic activity seep into the waters of Lake Kivu, which has high concentrations of dissolved carbon dioxide and methane. Generally the gases are contained in the deeper water under pressure – Lake Kivu is the world’s 18^th deepest lake, at 475 meters, more than 1,500 feet. But sometimes lakes experience overturns, with the deeper waters flipping to the surface. When gases are dissolved in the water and the pressure reduces, they can abruptly come out of solution like opening a carbonated beverage bottle. This happened catastrophically at Lake Nyos in Cameroon in 1986, asphyxiating 1700 people and thousands of cattle and other livestock. The possibility that Lake Kivu could do the same thing is a real threat to about two million people.

The critically endangered mountain gorilla lives in the Virunga Mountains, which also holds the research institute founded by Dian Fossey.

—Richard I. Gibson

December 16. East African Rift

We’ve been talking about the dismemberment of Pangaea and its biggest piece, Gondwana, for months now. The process is still going on, and the newest break within the old Gondwana continent is in its largest surviving portion, Africa. 

Map from Digital Tectonic Activity Map of the Earth (NASA)
with annotations by Gibson.

The East African Rift System is a present-day break that extends from the Dead Sea in Israel and Jordan, south through the Red Sea, separating Arabia from Africa, and into the African continent through Ethiopia, Kenya and Uganda, eastern Congo and Zambia, and into Mozambique and on offshore. All told, the system is more than 5,000 miles, 8,000 kilometers, long. It’s a big, complex break in the continental crust. 

As much as we have a pretty good handle on how rifting proceeds, our understanding of how and why such rifts begin is still pretty poor. There are multiple ideas for how the East African Rift started, ranging from some deep-seated mantle plumes, whose upwelling heat broke the crust apart, to crustal thickness variations that allowed magma to flow upward in some locations preferentially to others, initiating the rift process. Differences in crustal density might have the same effect as thickness variations. Whatever started the rift, it has since followed a pretty standard and expectable development.

Early in the process, around 30 million years ago, early Oligocene time, the upwelling magma breached the surface and flowed as extensive flood basalts in what are now Ethiopia, Somalia, Yemen, and adjacent areas. This point is called the Afar Triple Junction, because it is the focus for three branching rifts. To the northwest it’s the Red Sea, a young ocean basin where sea-floor spreading has just barely begun, to the northeast is the Gulf of Aden, true oceanic crust, and the mid-ocean ridge there continues into the Indian Ocean as the Carlsberg Ridge, the divide between the Indian tectonic plate and the African Plate. The third branch of the system extends from the Afar Triple Junction into the African Continent. 

Where the rift is in continental crust, in East Africa, the result is long narrow down-dropped troughs, called grabens. They are bounded by normal faults that have large offsets, many thousands of feet in some cases. The situation is very much like eastern North America must have been back in the Triassic as the Atlantic Ocean began to open. In Africa, it’s not one simple linear zone, but it curves and branches into two major segments on either side of Lake Victoria.

The fault-bounded troughs, the grabens, are obviously lower that the uplifted flanks, which tend to be mountainous, and the grabens or basins accumulate thick piles of sediment eroded off the mountains. In East Africa, the long, narrow lakes, such as Abaya in Ethiopia, Turkana in Kenya, Lakes Albert, Edward, and Kivu along the eastern border of Congo, Lake Tanganyika between Congo and Tanzania, Lake Rukwa, and Lake Malawi all lie in the down-faulted basins of the East African Rift. Lake Victoria, the second largest freshwater lake in the world, after Lake Superior, isn’t in a narrow fault basin, but it is related to the tectonic activity. It lies between the two big branches of the rift system, and formed when the uplifts to east or west dammed rivers flowing into the central basin. Victoria is a young lake, only about half a million years old or less, and it has dried up completely at least three times in its history, a reflection of changing climate conditions during the recent ice ages. Victoria is a shallow lake, less than 300 feet deep. In contrast, the deep troughs of the rift system hold some of the deepest lakes in the world. Lake Tanganyika, for example, reaches a maximum depth of more than 4,800 feet, and holds about 18% of all the fresh water on earth.

Volcanic activity continues in the region related to the rift process, including Mt. Kenya and Mt. Kilimanjaro, and the active volcanoes of Ethiopia and the Mountains of the Moon in Congo. Over time, the rifting has been sporadic. After the initial pulse of rifting and flood basalt eruption in the Oligocene, the faulting and real rift formation began in early Miocene time, around 22 million years ago or so. There was a period of several million years without too much going on, and the volcanic activity and earthquakes in the region today began in earnest about 5 million years ago or thereabouts.

The present rate of extension, about 6 millimeters per year, is slow, compared to typical oceanic crustal spreading rates, which are more like 20 millimeters a year, two centimeters. The slow rate is probably at least partially related to the strength and thickness of continental crust – it is more brittle, and harder to move than oceanic crust. But, at this rate, we should have a narrow ocean similar to the Red Sea separating the two parts of Africa by about 10 million years from now. The dismemberment of Gondwana continues, at least on this side. But to the north, Gondwana – India, Arabia, and North Africa, are more or less in a state of collision with Eurasia. All these things go on simultaneously.

* * *

Today we have the anniversaries of two significant earthquakes. On December 16, 1920, a quake hit Haiyuan County in Gansu Province, central China. The death toll estimate, 200,000, has been increased by modern estimates to almost 275,000, making it one of the most deadly earthquakes in human history. Its magnitude has been given variously from 7.8 to 8.5, but whatever it was, it shook the earth enough so that seiches – basically, a sloshing of the water in a relatively enclosed body, such as a lake – were recorded in the fjords of Norway. The location was along the Gansu thrust fault, a major fault where one of the continental blocks of North China is being pushed over the rocks to the south. Or maybe it’s better to think of it as the rocks to the south being pushed under the rocks of North China, because this quake is a result of the ongoing collision between India and Eurasia.

The second big earthquake on this day, December 16, was in 1811, at New Madrid, Missouri. It was the first of three quakes over a two-month period there that had magnitudes of about 7.5. They are among the largest historic earthquakes ever in North America. There were few deaths because the region was so sparsely populated, but the sequence of three quakes resulted in the formation of Reelfoot Lake in Tennessee, and the Mississippi River temporarily flowed backwards as a result of the forces. The December quake was strong enough to awaken people in New York City and to damage buildings in Cincinnati, Ohio. There’s a nice book chronicling these quakes, titled When The Mississippi Ran Backwards, by Jay Feldman (2012, Free Press Publishing).

—Richard I. Gibson

East African Rift 

Map from Digital Tectonic Activity Map of the Earth (NASA) with annotations by Gibson.

September 27. West Siberian Rift

Today I’m calling on your memory again, to think back on the vast flood basalts that erupted in Siberia just about at the end of the Permian. They’ve been implicated in the extinction event then, the most devastating extinction the earth has seen. 

Today, the basalt flows are exposed over extensive areas of Siberia, but west of those exposures, in the relative lowlands occupied by two huge river systems, the Yenisey and the Ob, the basalt flows are found in the subsurface of what’s called the West Siberian Basin. 

W. Siberian Rifts (green) from USGS (Ulmishek)

A few million years after the start of the Triassic and probably continuing for millions of years into the Triassic, a rift began to form in this West Siberian area. It was perhaps something like the rifts, the grabens, that were forming in eastern United States, but it was unlike them in that the whole system ultimately failed – no ocean was created by the rifting here, whereas the Atlantic formed from the rifting between North America and Africa. The West Siberian rifts were pretty big, though – one extends continuously for at least 1,800 kilometers.

Now visualize something like Nevada’s basin and range – long linear rift valleys adjacent to long linear mountain ranges. I’ve mentioned repeatedly that when you have such a situation, the erosion begins to fill those valleys – but this time, let’s focus on the mountains that are being eroded. The Permian and early Triassic flood basalts covered this region in vast, continuous sheets. When the rifting began, the sheets were broken – some segments were dropped down into the grabens, while other segments of the sheets were left on top of the mountain ranges, which are technically called horsts, a German word for heap or pile, something standing higher than the surrounding land.

The basalt on the tops of the mountain ranges got eroded away, leaving the basalt in the down-dropped basins. An alternative interpretation is that the basalts were erupting as the basins and ranges were forming, so the lavas flowed into the low-lying basins and were never deposited on the mountain tops, but I think the preferred interpretation is that the basalt flows were eroded off the mountains. In any case, what’s left, deep in the subsurface, is linear basins full of basalt and linear uplifts barren of basalt. The whole thing subsided even more, millions of years later, so all that complex structure, basins and ranges, is completely buried by later sediments, and the surface of the West Siberian Basin today is fundamentally a flat marshy plain.

Cross-section by Richard Gibson - source

So how do we know the basalt flows are down there? One way of course is to drill into the subsurface and get samples. But that’s expensive. Another way is to recognize that basalt is a dark, iron-rich rock, and a good bit of that iron is in the form of magnetite – an iron oxide mineral that is highly magnetic. We can measure the earth’s magnetic field at a distance from the magnetic rocks, from an airborne magnetometer, so that we can infer the distribution of magnetite-rich rocks in the subsurface. When Soviet geophysicists did that in the 1950s and 1960s, they revealed long linear magnetic highs – representing zones where there was a lot of magnetite – alternating with long linear magnetic lows, where magnetite was absent or less abundant. You’ve probably guessed that what they were defining were the grabens filled with basalt and the buried uplifts where the basalt had been eroded away.

So what? Well, those uplifts are buried beneath Jurassic and Cretaceous rocks that include some rich hydrocarbon source rocks and excellent reservoirs. The sedimentary rocks draped over the deep Triassic uplifts contain some of the largest natural gas fields in the world. And the gas fields coincide almost perfectly with the low values in the magnetic data, the uplifts where basalt is absent. Many trillions of cubic feet of natural gas from the West Siberian Basin have heated homes in Russia and much of Europe for decades. Back in 1990 I did an extensive analysis of the magnetic map of the Soviet Union for oil exploration, and in the process discovered this correlation between magnetic lows and gas fields. The Soviets had known it for decades, of course, but it was pretty satisfying to unravel the details of the relationships myself.

I have a link below to a nice paper from the USGS if you are interested in more about the oil and gas in the West Siberian Basin. Or email me.

—Richard I. Gibson

Links:
Siberian basalt flows 
West Siberian Basin Oil  (PDF - source of Map)
Cross-section by Richard I. Gibson
Urengoy gas field

September 8. Pangaea begins to break up

The other day, I said that in the Triassic, Pangaea was pretty much still intact as a single huge continent, with the big embayment on the east side, the Tethys Ocean. But I’ve also said there were hints of the great breakup that was about to begin.  

As the Cimmerian blocks began to rift away from the northeastern coast of Pangaea, the southern portion of the supercontinent, the old Gondwana, was probably rotating a bit, so that where Africa and Europe were attached, they began to pull apart at least to some extent, but it’s not really completely clear exactly what was going on there during the Triassic. Further north and west, through the complex mountain ranges that formed during the Caledonian, Alleghenian, and Appalachian Orogenies, over many millions of years, the compression due to collision was giving way to extension.

Triassic Globe by Ron Blakey, NAU Geology, under Creative Commons license (notes by Gibson)

Even as early as the late Permian, a rift, a pull-apart, had begun to form between what is now northeastern Greenland and northwestern Norway. That narrow strait might have allowed sea water to invade the basins of northern Germany and the Netherlands, where the Permian Zechstein salt formed. By the middle of the Triassic, the rifting seems to have extended a long way into the combined North America-European continent.

Tethys reconstruction globe from Stampfli & Borel 2002

How do we know this? There are extensive deposits of terrestrial sediments scattered through the region – mostly offshore today – from southern Greenland to west of Ireland and France and Iberia, and on the North America side, from east of Newfoundland around the margin to south of Nova Scotia – which was still attached to Africa, about where Morocco is today. This was not a complete seaway, but the rifting was making basins, similar perhaps to the basins that received the Old Red Sandstone back in the Devonian, after the first big mountains formed through this zone, the Caledonian Mountains. It was a complex array of zig-zagging rift basins, and I really think it’s fair to think of it like the East African Rift today – long, linear interconnected rift zones, in places with lakes, in places just lowlands receiving eroded debris off the adjacent highlands.

Let’s take a break for a minute and talk about rifts. When I say “rift,” I mean a major break in a continent, where two parts of the continent pull apart from each other. Ultimately, such a rift might become an ocean basin, with the two continental fragments bordering it on each side. This process is often driven by the generation of new oceanic crust at a mid-ocean ridge. Heat rising in convection currents from the deep mantle brings molten material to the surface – or at least near the surface – in a linear zone. As more and more such material rises, the previous material has to move out of the way – and the crust of the ocean spreads apart, away from the mid-ocean ridge. If that ridge started beneath a continent, the inexorable force of rising heat and magma will eventually break even thick continental crust. That’s what’s happening today in East Africa, and it’s what was beginning to happen during the Triassic where North America and Europe were attached. This is the birth of the modern Atlantic Ocean.

Rifting. (I have tried and failed to
determine the owner of this image;
if you are the copyright owner, please let me know.)

But during the early and middle Triassic, we didn’t have much in the way of open oceans yet. Probably just that narrow strait next to Greenland in the north, and possibly some ocean between North Africa and southern Europe. The rest was a diverse lowland, a sag, with linear mountains surrounding lake basins and continental river systems. Think of it like a big mass of cold caramel – soft enough to stretch some, but brittle enough to break eventually. As you pull the caramel apart, the middle will sag, and finally will break with a pretty sharp edge, assuming the consistency is just right. Since most of the rocks are under the Atlantic Ocean today, they are known only from remote sensing studies, including seismic data, and from wells drilled for oil and gas exploration. Not quite the same as having them exposed for geologists to take rock hammers to.

I think two big questions might be occurring to you at this point. First, what makes a rift start? And second, in this case, why did the rift run more or less along the zone where the original collision had created a huge mountain uplift that went from northern Greenland all the way to West Texas and probably beyond?

Oceanic rifts start where the linear edges of mantle convection currents rise toward the surface. The ultimate controls on the geometry, size, and position of convection currents are poorly understood – it’s the distribution of heat down in the mantle, and the complex response of the solid earth to that. And the solid earth is not uniform, so variety will be the name of the game. It’s also possible that some rifts begin because isolated mantle plumes, or hot spots like those at Yellowstone and Iceland, rise and weaken the crust, essentially encouraging the rift to radiate from that location. To break continental crust, much thicker and stronger than oceanic crust, probably depends on some special circumstances, such as a pre-existing state of stress, but it seems possible that a mantle plume might initiate such a continental break-up. This is still a controversial topic. Here's a 2014 paper on this idea, and see also this paper from 2014 for an opposing view.

As for the second question, why did the Atlantic Rift begin to form pretty much right along the zone where the continents had come together, one simple rationale is that such a zone, full of faults and inhomogeneities, would be the weak point in the system. The central cores of the continents – the cratons, which we outlined in January, and the word craton means “strong” – would have been much more resistant to breaking apart than the collision belt. You might argue that the collision zone made the crust even thicker, and with lots of igneous rocks and metamorphism, the suture zone, where the continents were welded together, ought to be the strongest part. Maybe it was. But it’s an observational fact that the break-up of Pangaea – at least between Europe and North America – followed the old collision, more or less. There are some interesting exceptions that we’ll talk about as the break-up proceeds over the next month or so.

* * *

Today’s birthday is Raphael Pumpelly, born September 8, 1837, in Oswego, New York. His geological work was wide-ranging, from Chinese coal fields to the copper country of Michigan, but he focused on economic geology of mineral deposits. He was the first to explore the Gobi Desert scientifically and he was also in charge of the Northern Transcontinental Survey of Dakota, Montana, and Washington Territories in the early 1880s. Pumpellyite, a low-grade metamorphic calcium-iron silicate mineral, was named for him.

—Richard I. Gibson

Tethys reconstruction globe from Stampfli & Borel 2002:    http://www-sst.unil.ch/research/plate_tecto/alp_tet_main.htm#Introduction 

Globe by Ron Blakey, NAU Geology, under Creative Commons license (notes by Gibson)


References: P.A. Ziegler, Evolution of the Arctic-North Atlantic and the Western Tethys, AAPG Memoir 43, 1988.

Mantle plumes cause rifts?